GlusterFS Georeplication gets really slow after a while

[MennaKhaled]

Gluster Version: 11

I started Geo replication a long while ago and it became really really slow.

So I decided to delete the volume at the destination, recreate it and start over.

Status at source:
root@sto01-glusterfs1-prd:/var/log/glusterfs/geo-replication/filesrepo01_192.168.16.67_filesrepo01# gluster volume geo-replication status

`_PRIMARY NODE      PRIMARY VOL    PRIMARY BRICK                       SECONDARY USER         SECONDARY                                               SECONDARY NODE    STATUS     CRAWL STATUS    LAST_SYNCED_
`-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
192.168.86.135    filesrepo01    /mnt/logicalvolume01/filesrepo01    glusterfsreplicator    ssh://glusterfsreplicator@192.168.16.67::filesrepo01                      Passive    N/A             N/A
192.168.86.137    filesrepo01    /mnt/logicalvolume01/filesrepo01    glusterfsreplicator    ssh://glusterfsreplicator@192.168.16.67::filesrepo01                      Active     Hybrid Crawl    N/A
192.168.86.136    filesrepo01    /mnt/logicalvolume01/filesrepo01    glusterfsreplicator    ssh://glusterfsreplicator@192.168.16.67::filesrepo01                      Passive    N/A             N/A

Again at the very beginning the rate is relatively high at 300+ GB/day but now it's almost 200 MB/day. What causes this behavior. The source is a changing volume but that should be acceptable, correct?
what would be a good workaround? I have 7T of data.
Would getting everything to destination using any other tool, and then starting the georeplication to keep the deltas make sense?